Cerebral vasospasm remains the leading cause of morbidity and mortality after aneurysmal subarachnoid hemorrhage (aSAH). Predicting its occurrence in a timely manner for therapy intervention is critically important in improving outcome after aSAH but remains unsatisfactory using existing techniques. The overall goal of this project is to develop and validate a novel data fusion algorithm for predicting vasospasm after aSAH that only requires bedside measurements of arterial blood pressure (ABP), intracranial pressure (ICP) and cerebral blood flow velocity (CBFV). The specific objectives of this project are 1) To develop a vasospasm warning generation algorithm based on the lumped proximal (n) and distal (r2) cerebral arterial radii estimated from the data fusion process; 2) To compare the data fusion approach with the existing Transcranial Doppler (TCD)-based vasospasm diagnostic criteria; 3) To compare the data fusion approach with model-independent methods for vasospasm prediction. A constrained nonlinear Kalman Filter is applied in the data fusion process to estimate the lumped proximal (n) and distal (r2) cerebral arterial radii that are two internal state variables of an intracranial pressure dynamic model. It is hypothesized that development of vasospasm will show a consistent trend in the estimated n and/ or r2 and that this trend is detectable by a statistical trend detection algorithm. The detection of such a trend will then fire a warning of impending vasospasm. It was estimated that annual number of aSAH patients is about 30,000 in US alone, among which up to 70% can develop angiographic vasospasm with possible vasospasm in small cerebral arteries in remaining cases. Cerebral ischemia due to vasospasm not treated in a timely fashion can lead to devastating outcome. A promising way to improve outcome after aSAH is to detect it even before angiographic evidence and act with interventions. If validated, the proposed data fusion vasospasm assessment method could achieve this goal in a low cost and in a compatible way with the current clinical practice, a desirable feature to gain wide clinical acceptance of a new approach. [unreadable] [unreadable] [unreadable]